Biomineralization in seeds: developmental trends in isotopic signatures of hackberry

PAL/ 0 ELSEVIER

Palaeogeography, Palaeoclimatology, Palaeoecology 138 (1998) 259 269

Biomineralization in seeds" developmental trends in isotopic signatures of hackberry A. Hope Jahren a,,, Mark L. Gabel b, Ronald Amundson c a Earth andAtmospheric Science, Georgia Institute ~?fTechnology, Atlanta, GA 30332-0340, USA b Department of Biology, Black Hills State University, Spearfish, SD 57799-9035, USA Division of Ecosystem Science, University of CaliJbrnia at Berkeley, Berkeley, CA 94720, USA Received 27 January 1997; received in revised form 22 August 1997; accepted 9 September 1997

Abstract Experiments on hackberry (Celtis) fruits collected over the course of a growing season at two localities were undertaken to assess the isotopic composition of the intricate mineral (CaCO3 and SiO2" nil2 O) and organic structure of hackberry endocarps. Hackberry fruits contain biomineralized aragonite-rich endocarps and have been documented as fossils in a variety of sites in North America and Eastern Europe ranging in age from historical to millions of years old, giving them potential as paleoclimate indicators. In the modern hackberries studied, three distinct stages of development were observed: stage 1 (0-55 days) shows a rapid increase in the whole mass of the hackberry drupe as basic components differentiate; stage 2 (day 55-85) involves a short (30-day) period of slow growth overall, but is concurrent with a sharp rise in the mass of calcium carbonate mineral; stage 3 (day 85 to full maturity) involves a sharp increase in total mass due to mesocarp development. Fruit tissues were found to have lower i513C values than stem tissues at both sites of collection; there was very little difference between the ~5180 values in the endocarp carbonate at the two sites. The most variable 613C and 6180 values of endocarp were observed in stage 1 development. After stage 2 begins, the ~180 values in endocarp carbonate stabilized. The 613C value of opal-occluded organic carbon appeared to be constant throughout the growing season. © 1998 Elsevier Science B.V.

Keywords." biomineralization; paleoclimate; stable isotope; phytolith; fossil seed

I. Introduction The hackberry (Celtis) is a deciduous tree in the elm family (Ulmaceae). This irregularly shaped tree sometimes occurs as a shrub and has a wide, but fragmented, distribution t h r o u g h o u t N o r t h America and Eastern Europe, including Anatolia (Turkey). The hackberry is viable in a wide range * Corresponding author. Fax: + 1 404 894 5638. E-mail: [email protected] 0031-0182/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S0031-0182(97)00122-3

o f habitats, including deciduous riparian w o o d lands, various shrub communities, in rocky ravines, and as scattered individuals in grasslands (Ginski, 1977; Plummer, 1977; Carmichael et al., 1978; Brown, 1982; Albee et al., 1988; DeBolt, 1992), and is able to grow at elevations ranging from 200 to 2000 m above sea level (Elias, 1980). Celtis produces a fleshy, globose drupe containing a single seed. Fruits ripen in late fall and are dispersed by birds, rodents, other small mammals, and gravity. Occasionally, the seed has been

260

-t.tt. YahrvH cl a/.

Pa/aeovco,k, raldll', t~a/acoc/imago/oW, Palacocco/o~)' 13~¥ f 199,¥/ 259 269

observed to remain on the tree until the following spring ( Lamb, 1915 ). The drupe contains a fleshy mesocarp and a biomineralized aragonite-rich endocarp, all of which encase the endosperm. The durability of the 'biomineralized' aragonite endocarp contributes to its preservation in the l\~ssil record: fossil Cehis endocarps are found throughout the Quaternary sediments of the Great Plains, and are common botanical remnants in sediments of the Miocene and Oligocene epochs. Because the Cehis fruit is edible and was a favorite condiment of some American Indian groups, fossilized ('elti,~ endocarps are commonly found in Great Plains Holocene archeological sites (Thomasson, 1991 ). Fossil hackberry endocarps have been excawtted from a variety of sites ranging in age from historical to millions of years old, but at present there is very little information on their structure, isotopic composition and relation to local climate. In this paper, we report on our investigation o1" the morphological, mineralogical, and isotopic development in hackberry fruits collected from two locations in the Northern Great Plains during the growing season of 1993. The motivation of this research was our interest in the suitability of the stable isotope composition of hackberry endocarps as paleoclimatic indicators (Jahren, 1996). Elsewhere, we have researched the durability of the endocarp through time in terms of its mineralogy and 14C composition (Wang et al., 1997): this work suggests that the general composition of endocarps is well-preserved, at least through the Quaternary.

2. Materials and methods

2.1. Fie/d colleclion Cehis occidentalis fruits were collected from two planted trees in Austin, Minnesota and one in Spearfish, South Dakota during the growing season of 1993. Samples were gathered over a period of 18 weeks beginning with the first appearance of fruit, and continuing until the fruits had reached maturity. Each collection consisted ot" picking 10 20 fruits from each tree. Following collection, specimens were air-dried for several

weeks and then weighed. One fruit per tree was analyzed isotopically for each collection, and at least three replications were made of each endocarp, with less than 0.2%,, variability between replications. 2.2. ('hemica/ ana/rses

The mass of total carbonate in fruit was determined by reacting a known mass of ground sample with 100% phosphoric acid at 25 C, cryogenically purifying the CO, produced, and determining its quantity manometrically (McCrea, 1950). Components of drupes (meso +exocarps, endosperm) were either separated physically, through cutting and scraping, or chemically, with the following chemical treatments: whole endocarp was treated chemically with 35% H2Oz for 24 h at room temperature in order to remove oxidizable organic materials, then acidified with 6 M HCI for 12 h in order to remove all carbonate mineral. Persisting endocarp remains were exposed to 60% HF for 2 weeks in order to completely dissolve minerals, leaving a network of occluded organics. Cellulose was isolated from plant tissue using a method adapted from Harborne (1984) in which plant samples were boiled in H20, then bleached with sodium chlorite and acetic acid to remove lignin, and rinsed. Hemicellulose was removed with 17g, NaOH: and further rinsed. The remaining pure cellulose was lyophilized and stored. Whole hackberry drupes were impregnated under vacuum in Epo-tek@ 301-2 optically transparent epoxy, then thin-sectioned to a thickness of 30 lain according to standard petrographic techniques. Slides were then examined under a Nikon Optiphot-2 petrographic microscope with camera attachment, Endocarp fragments examined via scanning electron microscope (SEM) were mounted on aluminum stubs and coated with Au. All SEM work was done on a IDIS microscope (stage 2) at 10 kV. Magnification levels ranged from 27 to 580 ×, with most features showing best resolution at about 400 x. Selected samples were subjected to X-ray diffraction ( X R D ) analysis in order to ascertain mineralogy, and to Fourier Transform Infra-Red

A.H. Jahren et al. / Pa/aeogeography, Palaeoclimatology, Palaeoecology 138 (1998) 259 269

(FTIR) analysis in order to characterize organic components. Samples were introduced to the FTIR instrument as a 1 2% concentration in KBr. Organic carbon samples were prepared for stable isotope analysis via combustion in sealed vycor tubes containing Cu, CuO, and Ag (Minagawa et al., 1984). The released CO2 was purified cryogenically, measured manometrically, and collected for 13C/12C measurement on a stable isotope ratio mass spectrometer. Carbonate samples were prepared for stable isotope analysis through a pretreatment in bleach designed to minimize the effects of organic carbon upon carbonate reaction (Jahren, 1996) and then by reaction with 100% phosphoric acid (McCrea, 1950) in separate reaction vessels (as described above) or at 90°C in an automated 'common acid bath' carbonate device. Measurement of stable isotope ratios of the liberated CO2 in both methods was performed on a VG Prism stable isotope mass spectrometer at the Lawrence Berkeley National Laboratory. All isotope values are reported in the delta notation:

261

with 6 M HC1 (for 12 h in order to remove all carbonate mineral), a persisting endocarp remains (Fig. 4), consisting of a skeletal framework, or 'scaffolding', upon which the carbonate mineral was overlaid. Combustion analysis indicated that this 'scaffolding' is ~15% organic carbon by weight, indicating that it consisted predominantly of minerals insoluble in HC1. The remaining skeletal framework was exposed to 60% HF for 2 weeks in order to completely dissolve minerals, leaving a network of occluded organics (Fig. 5).

3.2. Characterization of mineral components

where the standard is VSMOW for R = 180/160, and the standard is VPDB for R = 13C/12C.

X-ray diffraction of bulk endocarp indicates that endocarps consist primarily of aragonite (CaCO3) and that the acid-resistant mineral 'skeleton' consists of opal (SiO2"nH20). Cystoliths in the Acanthacea plant Beloperone caliJornica have also been found to be composed of both calcium carbonate and silica (Hiltz and Pobeguin, 1949), and this deposition of both calcium carbonate and silica, apparently in a single cell, is thought to be most unusual (Simkiss and Wilbur, 1989). We find interlaced amorphous silica and aragonite precipitated in the biomineralized endocarp of the hackberry in most specimens, but occasionally the opal is not present, for unknown reasons.

3. Results and analyses

3.3. Characterization of organic components

6=

Rsample -- Rstandard

× 1000 [%,,]

Rstandard

3.1. Light microscopy and SEM results The mesocarp, the mineral endocarp, and the endosperm are illustrated in cross-section under light microscopy (Fig. 1). Selective chemical treatments of the mineral endocarp (physically separated from mesocarp and endosperm) reveal an intricate structure. Whole endocarp (treated chemically with 35% H202 for 24 h at room temperature in order to remove oxidizable organic materials) is illustrated in Fig. 2. This SEM image clearly reveals ~30 ~tm diameter, 'honeycomb' mineral units of the endocarps that are more visible at higher magnification (Fig. 3). Upon acidification

Fourier Transform Infrared Spectroscopy (FTIR) was used to characterize organic compounds present in different components of the hackberry fruit. Results are presented in Table 1. Results indicate the presence of three compounds in all components: RCH=CH2 (alkene), fatty acid or ester, and alkane. In addition to these compounds, the mesocarp and opal-skeleton occlusion contain additional alkane and alcohol compounds, while the mesocarp only contains ether compound. The fact that several compounds are found in all components suggests that the organic chemistry is rather homogeneous throughout the fruit, but that cellulose and lignin are not found in the endocarp.

262

.4. H. Jahren el a/.

Pa/aeo~,,eographv. Palaeoc/mmto/o~v, F'a/aeoecolo,t:v 138 I 199~'f) 259 269

4 +

0

mesoearp Q

o

endosperm

-0 0

Q J

1 mm

'11"'~

•

I

*I Fig. I. Photograph of hackberr 5 m cross-section.

Fig. 2. SEM image of hackberry endocarp that has been treated with 35% HzO 2 in order to remove oxidizable organics. × 166.

3.4.

Fig. 3. ttigh-magnification SEM image of hackberry endocarp that has been treated with 35% H_,O=in order to remove oxidizable organics: here the honeycomb mineral units are distinctly ~isible. × 580.

Thin-section petrography results

D e v e l o p m e n t o f the m i n e r a l e n d o c a r p is clearly a p p a r e n t in a time-series c o l l e c t i o n o f p h o t o g r a p h s

taken under the petrographic microscope ( F i g s . 6 - 9 ) . E x t e n s i v e d e v e l o p m e n t in m i n e r a l e n d o c a r p s was o b s e r v e d to o c c u r b e t w e e n 74 a n d

A.H. Jahren et al. / Palaeogeography, Palaeoclimatology, Palaeoecology 138 (1998) 259-.269

263

fleshy mesocarp). Components were isolated using the methods described earlier in this section, and the results of the isotopic analyses are presented in Table 2. The stability of the isotopic composition of the endocarp throughout the growing season is depicted in Figs. 10-12. Fig. 10 plots the fi13C values of endocarp carbonate against time; Fig. 11 plots the fi180 values of endocarp carbonate against time; and Fig. 12 plots the 613C of organic carbon occluded within the endocarp's opal skeletal structure against time.

Fig. 4. SEM image of hackberry endocarp after treatment with both 35% HzO 2 and 6 M HC1. Only the skeletal framework of the honeycomb mineral structure remains. × 424.

Fig. 5. SEM image of hackberry endocarp after treatment with 35% H202, 6 M HC1 and 60% HF. Only the organic material occluded within the skeletal framework of the honeycomb mineral structure remains. × 3030.

133 days after appearance, both with respect to added mineral mass, and definition of mineral structure.

3.5. Stable isotope analysis results Selected portions of the hackberry tissue collected were analyzed for carbon and oxygen stable isotope content: (stems, stem cellulose, organic carbon occluded within the opal skeletal structure of the endocarp, endocarp carbonate, and the

4. Discussion and conclusions

Williamson and Coston (1989) have described a three-stage growth pattern in the development of peach fruit as the following: stage 1, rapid cell division and increase in cell size in the pericarp; stage 2, a short period of slow pericarp growth; and stage 3, enlargement of the mesocarp cells. This general 3-stage model of fruit development has been confirmed in peaches, as well as other fruits by many other workers, including Ognjanov et al. (1995), and Luthra et al. (1980). Three distinct stages of development were also observed in the hackberry at our two sites. Fig. 13 shows mass of hackberry drupe plotted against days after its first appearance. The various 'stages' of development have been labeled according to patterns observed in the data and in the morphology of the hackberry drupes. Stage 1 (0-55 days) shows a rapid increase in the whole mass of the hackberry drupe as the basic components begin to differentiate themselves. Stage 2 (day 55-85) involves a short (30-day) period of slow growth overall, but it is concurrent with a sharp rise in the mass of calcium carbonate mineral in the drupe. The final stage (3) involves a sharp increase in total mass as the mesocarp acquires most of its mass. Three-stage development in the hackberry drupe has implications for the timing of biomineralization in the organism. Fig. 14 plots percent calcium carbonate (by mass) against day after first appearance of the drupe. The striking feature of Fig. 14 is that during stage 2 the total mass of the drupe is found to be up to 20% carbonate in composition

264

4 . t l J a h r e n el al,

Palaeot,,eograph ', P I~ e ~ '/in llolov.l. Pah e o e c o l o ~ ' 138 : 1998) 259 269

Table 1 Results of FTIR analysis of the three mare morphological components of hackberry fruit" Hackberry component

Signilicant peak absorption {cnl / }

Class of compounds present

Interpretation

Mesocarp

11156 1645 1745 2854 2925 ~337

( (' C (" ( O

O C 0 H H H

ether RCH CH 2 (alkene) fany acid or ester alkane alkane alcohol

655 745

C C C O C 1t

R C H = C H 2 (alkene) fatty acid or ester alkane

I100 1642 1736 2918 3346

(" O (" (' ( 0 (' tt O Il

R , C H O H (alcohol) RCH -CHe(alkene) IEtty acid or ester alkane alcohol

Eudosperm

Opal-skeleton occluded organics

%ample was a mature fruit sample from Austin, MN (1993).

Fig. 6. Photograph of hackberry thin-section: taken under the petrographic microscope with crossed potars and quartz plate inserted. Scale: the long axis of the i m a g e - I mm. The endocarp is shown at 41 days after appearance, and lacks an k visible pattern in the mineral component.

Fig. 7. Photograph of hackberry thin-sectiom taken under the petrographic microscope with crossed polars and quartz plate inserted. Scale: the long axis of the image - I ram. The endocarp is shown at 74 days alter appearance, small, honeycomb-shaped mineral units become apparent.

(during stages 1 and 3 the total mass contains a much smaller % carbonate). This pattern is apparent in petrographic thin sections ( Figs. 6 9) where most of the mineral mass of the endocarp appears to have been added during a short period in the middle of drupe development. The pattern of organic material addition/mineral material addition/organic material addition is intriguing from a plant physiological perspective. Many

studies clearly illustrate the ion-specific nature of element translocation in plants (Sale and Campbell, 1980; Dornbos and McDonald, 1986: lskander, 1987: Laszlo, 1990), but with respect to the hackberry, there is a clear timing of selective element transport and deposition. During fruit development, C-substrates travel from their synthesis sites in the vegetative parts of the plant through the phloem to the site of the developing

A.H. Jahren et al. /' Palaeogeography, Palaeoclimatology, Palaeoecology 138 (1998) 259 269

+1

+1

+S

+S

+1

¥1

265

©

~S

r~ t,e5 e,¢5,

Fig. 8. Photograph of hackberry thin-section: taken under the petrographic microscope with crossed polars and quartz plate inserted. Scale: the long axis of the image = l ram. The endocarp is shown at 133 days after appearance; the mineral endocarp has undergone extensive development both in added mineral mass, and definition of mineral structure.

÷

"4"

+1 IS

+1

I S

+

÷1

~S

I S

_--.

.~. ~,

÷

~

II

4

+1 --

o

~

¢'1

IS

J S

+

~, g

+

-~l ~

b

+1 Fig. 9. Photograph of hackberry thin-section; taken under the petrographic microscope with crossed polars and quartz plate inserted. Scale: the long axis of the image = 1 mm. The endocarp is shown at 150 days after appearance: the mineral portion of the endocarp has apparently stabilized, and the periphery has become more defined in structure.

fruit (Thorne, 1985; Patrick, 1990). Studies have shown that phloem-mobile elements, such as K+, Mg 2+, C and N can also be re-translocated from stem and leaf tissue to the developing fruit (Pitman, 1975; Mauk and Nood6n, 1992), and t h a t the ' s i g n a l ' f o r this r e t r a n s l o c a t i o n m o b i l i z a t i o n is h o r m o n a l ( N e u m a n a n d N o o d 6 n , 1983: N o o d d n , 1987). C a r b o n in h a c k b e r r i e s is u l t i m a t e l y d e r i v e d f r o m the a t m o s p h e r e as the p l a n t fixes CO2 i n t o o r g a n i c

+1 I S

'-'2

÷1

~,

S

e~

=

©

",2,

< 2 o o=

o

~

~

E

~

-u N

A.H. JahreH cl a/.

266

Palaco.,e Lk~ral~t(I', Pa/acoc/m;alo/ok, r Pa/aeoeco/oe ' 138 (1998) 259 26?

-II __

±

ca

21111 Austin, MN Ilackberfics j Spearfish. SD ttackhcrrtcs I

•

-12

-

.

I 75

-13

is0

-

~= - i 4

,,,~,~,,I,~,.,,*asa,oF~. Stage 2 (a(o m tlee #~1)0drupe

~

I •

-is

"=

Oq

-"

2 It ~(~ .O

/

Stage I

I5

1 '-~

-17

7c

. •

e" m

-16

2~

Stage 3 •

,.,, ....... .,~0~,,,<,w

25

r

m,,,,,,t ,,,-~ ,7tNIdrc,r*

I

l0

;

-18 =

-19

i) • 20

-21 4 0

6 0 8 0 days a f t e r a p p e a r a n c e

I (lO of d r u p e

121) t1993)

I 41)

Fig. Ill. Plot showing change in 6' 3(, xaluc of cndocarp carbon ate with time. 31 Austin. MN ttackberrics Spcarllsh, SlY llackbt'rrics

•

5

•

-20 20

•

/' 25

•

i

•

.

.

4 (I 6 0 da3,s after appearance

.

.

80 of drupe

0 12 0

.

I00 (11)93i

t I~. I 3. Plot showing change in mass of hackberr 3, drupe with time. For Ihis figure, hackberries were collected from two trees, u7111 ant #805 in Austin. Minnesota during the growing season of IcJ93. Mass (air-dry) of the entire drupe is shown in open ~ymbols and is scaled against (he left ~-axis. Mass of calcium czlrbonate mineral i n I h e drupe ~s s h o ~ v t l ill closed symbols a n d :> scalcd agalns: tile fight I'-£tXIS. The dotted line is a simple interpolation from mass to mass m the uee a700 whole drupe I l l e i l s U r e l l l e l l Is.

6 ea

,.~

29 ¸

28

tee #7(~1 dnl~" . . . . . . . tree #805 drut~

StaRe 3

Staee 2 20 ~-

27 i

[

26 ! . 2 0

• .

. 4 0

S'ta£,e ]

.

. 6 0

days a f t e r

.

. 8 0

10 0 of d r u p e

appearance

1211 (1993)

I40

g I 0

Fig. 11. Plot showing change in 6 ~ O ~ahle of endocarp carbonate with time.

;°°,-"

-24~

~z

i -25 [

"~- - 2 6

•

li

,\uslin, MN Hackbciliex Sl~.'arlish, SD Ilackherlics

-27

A

e

-29

-~ - 3 o [ 7.¢

•

-31 20

(I

4 I)

6 0

days a f t e r

i 8 0 appearance

I O0 of d r u p e

~

after

II

appearance

~i I! of d r u p e

I Oil 19931

I 21)

f.ig 14. ['lol showing change m percent calcium carbonate ~bx massl m whole h a c k b e r l ) drupe with time. All hackberries in lhis plol ,acre collected in Austill. MN during the growing season of 1993.

i

-28 :

4

days

1

"~

•

2 n

I 20

140

<1993i

Fig. 12. Plot showing change in 61~(" ~alue of organic carbon occluded within the opal-skeletal structure of the endocarp with time.

c a r b o n c o m p o u n d s These organic c a r b o n c o m p o u n d s are then either t r a n s p o r t e d to a site o f tissue d e v e l o p m e n t as an incomplete sugar, or are respired back to CO2. Because C O 2 is a relatively soluble gas, respired CO2 can be "stored' by solubilizing it into plant tissue waters. A f t e r it has been t r a n s l o c a t e d and c o n c e n t r a t e d in subcellular comp a r t m e n t s , dissolved C O , acts as a means to

A.H. Jahren et al. /Palaeogeography, Palaeoclimatologv, Palaeoecology 138 (1998)259-269

mediate cellular pH, as supply of readily available protons and minor reorganization creates carbonic acid (Simkiss and Wilbur, 1989). The other important component of mineral material addition is the incorporation of Ca 2+ into the plant. The ultimate source of Ca 2+ in all plants is dissolved calcium in soil water. It is unlikely that gross uptake of Ca 2+ by the plant is a limiting factor for fruit development, based upon the quantification of standard metabolic Ca 2+ uptake rates inferred from Sr 2+ tracing studies (Laszlo, 1994). Interestingly, both xylem flow rates and the concentration of Ca 2+ in xylem fluid have been shown to fluctuate substantially during fruit development (Nood6n and Mauk, 1987), suggesting that supply to developing fruit tissue may be a major sink for xylem-dissolved Ca z+. Patterns of Ca 2+ demand may even be genetically informed: P. vulgaris, fruit Ca 2+ is supplied predominantly by redistribution from the pod (Mix and Marschner, 1976), suggesting that translocation from storage tissue may be a major source of fruit Ca 2+. Another important factor is that Ca 2+ is phloem-immobile, and must be transported via a different route, such as the xylem or other non-vascular routes (Pate, 1975; Pate and Hocking, 1978; van Bel, 1990). The working hypothesis for the development of the hackberry fruit that we wish to put forward as a result of these observations includes the following: (1) development of early season structures - during this phase of development preliminary fruit structures are constructed from phloemsupplied organic carbon compounds. Components, such as the fruit stem, the basic definition of the pericarp, and the organic matrix for subsequent mineral deposition are constructed. This phase is equivalent to developmental stage 1 described by Williamson and Coston (1989); (2) at about 55 days after first appearance, the hackberry drupe enters the second phase of its development during which the bulk of the mineral endocarp forms and other late-season structures, such as the mesocarp and the endosperm, add substantial mass. The switch from stage 1 to stage 2 development is probably triggered hormonally, and the most important driver of this hormonal switch is probably a large increase of Ca 2+ in the xylem and in cell walls due to either translocation from storage

267

tissues in stems and leaves or increased Ca 2+ uptake by the plant. Mineral endocarp development continues and gradually plateaus into the fully developed endocarp; (3) the mesocarp continues to develop until senescence. The different phases of development in the hackberry are expressed in the way that 13C is partitioned into the various components of hackberry tissue. Table 2 shows that fruit organic structures have more negative 813C values than stem organic structures at both field sites. This may be the result of the photosynthetic capability of 'green' fruit tissue: photosynthesizing structures have previously been found to be 2 4%o more negative in 613C values than non-photosynthesizing tissues in the same plant (O'Leary, 1988). Table 2 also illustrates the systematic difference between the stable isotope values at Spearfish, SD and Austin, MN ("Spearfish Austin" row, Table 2). Specifically, stem tissue 613C values in Spearfish, SD plant tissue tend to be ~0.60%,, greater than 8~3C comparable structures from Austin, MN. In fruit tissue, the 813C isotopic differences between the two sites increases to 1.0%0 or more in comparable structures. Growing season temperatures in Austin, MN are very stable and unchanging during the months of June, July and August, with an average daily value of 22.1°C. In contrast, during the month of June in Spearfish, SD average daily temperatures only reach 17. t°C; then in July and August, average daily temperatures increase to 23.1°C. All reported temperatures are 30-year averages of 1960-1990 data (Bair, 1992). It is probably this increase in temperature during the late season that drives the increase in 813C isotopic difference between similar structures at the two sites. It is interesting to note that there is very little difference between the 8~sO values in the endocarp carbonate of hackberries at the two sites. As many workers have noted, the 81so of calcium carbonate is a strong function of the 81sO of the equilibrium solution and temperature (McCrea, 1950). Temperature of precipitation has been shown to affect the 81sO of resultant calcium carbonate ( Urey, 1947), but in this system, the differing temperatures of the two sites have not made the 81sO in endocarp carbonate strongly different: however, it is likely that the temperature

268

4.11, Jallrcll cta/.

I~a/ocogcu,wapllt, I~alucoc/inlal,flo:,,,r. Pa/aeoccolog)' 13,~' (1998) 259 269

Table 3 Estimates of within-site total variabilities based upon the data o1" Figs. 10 12 Component of hackberr3

Variahilil3 through gro~ing scason ('%~1

Variability in late development phase onl) {%,,)

6~3C in endocarp carbonate 31sO hi endocarp carbonate a'-~c in opal-occluded organic C

S.o ~,.() 0.~

3.0 1.2 0.9

effect has been offset by differences in the 6t~() of tree source-water at the two sites, as well as differences m relative humidity which determined water use efficiency. In order to pursue these speculations, we are presently completing studies designed to evaluate possible correlations between 61sO values in hackberry endocarp carbonate, the ~1~O value of site environmental water and site growing season temperature (Jahren et al.. submitted). Despite the fact that the ~?~C difference between similar organic components changes by about +0.5%,, from early to late season at the two sites. organic components are roughly equally different from endocarp carbonate 6~3C at both sites throughout the season, as is shown by a comparison of mesocarp organic C and endocarp CaCO, columns in Table 2. One of the goals of this study was to assess the variability of stable isotope composition in hackberry structures during the course of the growing season, in order to evaluate fossil hackberry endocarps as possible paleoclimate indicators. For example, the 6~3C value of endocarp carbonate was observed to fluctuate greatly over the coursc of the growing season (Fig. 10). The majority of this fluctuation, however, is encompassed by the stage 1 development and represents early stages of carbonate precipitation and the possible use of translocated metabolized C. The trend in variability of 8t80 in endocarp carbonate is similar, but on a more restricted scale ( Fig. 11 ). After stage I development ends. and the phase involving substantial carbonate precipitation begins, the value of ~tso in endocarp carbonate has stabilized greatly. In contrast, the value of /513(7 in opaloccluded organic carbon appears to be stable throughout the growing season (Fig. 12). From these results, the within-site variabilities

are estimated as presented in Table 3, and indicate that 6~3C m endocarp carbonate is much more wmable than 6180 in endocarp carbonate, and that variability of both signatures decreases significantly with fruit maturity. Therefore, when choosing hackberry endocarps l\~r analysis in a paleoclimate context, it is clearly best to select against obviously immature, largely unformed 'stage 1' endocarps when possible. Furthermore, in evaluating the variability in terms of potential uncertainty in a paleoclimate reconstruction, it is important to remember that the numbers expressed in Table 3 represent both growmg season variability and within tree variability, since sampling over the growing season was necessarily done on different fruits at each stage in development.

Acknowledgements The authors are grateful to Mr. C.E. Jahren l\~r collecting hackberries and maintaining the study site in Austin, MN, and to Dr. B. Herbert for pert\~rming FTIR analyses. This work was supported in part by an NSF Graduate Fellowship to A.H,J., and an NSF Geologic Record of Global Change Grant to R.A.

References Albec. B.,I., L.M. Schultz, S. Goodrich, 1988. Atlas of tile Vascular Plants of Utah. Utah Museum of Natural History, Occasional Publ. No. 7. Salt Lake City, UT, 67(1 pp. Bair. F.E. (Ed.), 1992. The Weather Almanac. 61h ed. (}ale Research. Brown, D.E., 1982. Great Basin montane scrubland. In: Brown. D.E. (Ed.). Biotic Communities of the American Southwest United States and Mexico. Desert Plants 4 (1 4), pp. 83 84.

A.H. Jahren et al. / Palaeogeoeraphy, Palaeoclimatology, Palaeoecoh)gy 138 (1998) 259 269 Carmichael, R.S., Knipe, O.D., Pase, C.P., Brady, W.W., 1978. Arizona chaparral: plant associations and ecology, USDA Res. Paper RM-202, Fort Collins, CO, 16 pp. DeBolt, A.M., 1992. The ecology of Celtis reticulata Torr. (netleaf hackberry) in Idaho, M.S. thesis, Oregon State University, 161 pp. Dornbos, D.L., McDonald, M.B., 1986. Mass and composition of developing soybean seeds at five reproductive growth stages. Crop. Sci. 26, 624 630. Elias, T.S., 1980. In: The Complete Trees of North America, Times Mirror Magazines, New York, NY, 948 pp. Ginski, R.L., 1977. Regeneration and distribution of sycamore and cottonwood trees along Sonoita Creek, Santa Cruz County, Arizona. In: Johnson, R.R. Dale, A. (Tech. Coord.), Importance, Preservation, and Management of Riparian Habitat Proceedings, USDA Gen. Tech. Report RM-43, Fort Collins, CO, pp. 116 123. Harborne, J.B., 1984. Phytochemical Methods A Guide to Modern Techniques of Plant Analysis, 2nd ed. Chapman and Hall, London, 288 pp. Hiltz, P., Pobeguin, T., 1949. On the constitution of cystoliths of Ficus elastica. C.R. Hebd. Seances. Acad. Sci. 228, 1049 1051. Iskander, F.Y., 1987. Maturation stage and mineral content in soybeans, Food Chem. 24, 29-35. Jahren, A.H., 1996. The stable isotope composition of the Hackberry (Celtis) and its use as a paleoclimate indicator. Ph.D. thesis, University of California at Berkeley, 136 pp. Jahren, A.H., Amundson, R. Kendall, C., Correlation in oxygen istopic signatures of North American modern hackberry endocarps and environmental water. Quat. Res., submitted. Lamb, G.N., 1915. A Calendar of the Leafing, Flowering, and Seeding of the Common Trees of the Eastern United States. Weather Bureau, U.S. Monthly Weather Review, Suppl. 2, Part I, 19 pp. Laszlo, J.A., 1990. Mineral contents of soybean seed coats and embryos during development. J. Plant Nutr. 13, 231 248. Laszlo, J.A.. 1994. Changes in soybean fruit Ca 2+ (St 2+) and K + {Rb+) transport ability during development. Plant Physiol. 104, 937-944. Luthra, R., Sharma, N.B., Paul, Y., 1980. Comparative study of oils of seed kernels of some Prunus species. Plant Biochem. J. 7. 159-166. Mauk, C.S., Nood6n, L.D., 1992. Regulation of mineral redistribution in pod-bearing soybean explants. J. Exp. Bot. 43, 1429 1440. McCrea, J.M., 1950. On the isotopic chemistry of carbonates and a paleotemperature scale. J. Chem. Phys. 18, 849 857. Minagawa, M., Winter, D.A., Kaplan, 1.R., 1984. Comparison of Kjehldahl and combustion tube methods for measurements of nitrogen isotope ratios in organic matter. Anal. Chem. 56, 1859 1861. Mix, GP., Marschner, H.R., 1976. Redistribution of calcium in bean fruits during seed development. J. Plant Physiol. 80, 354 366.

269

Neuman, P.M., Nood6n, L.D., 1983. Interaction of mineral and cytokinin supply in control of leaf senescence and seed growth in soybean explants. J. Plant Nutr. 6, 735 742. Nood+n, L.D., 1987. Soybean seed growth: regulatory mechanisms in the whole plant. In: Newman, D.W., Wilson, K.G. (Eds.), Models in Plant Physiology and Biochemistry, Vol. lI. CRC Press, Boca Raton, FL, pp. 145 147. Nooden, L.D., Mauk, C.S., 1987. Changes in the mineral composition of soybean xylem sap during monocarpic senescence and alterations by depodding. Physiol. Plant 70, 735 742. Ognjanov, V., Vujanic-Varga, D., Misic, P.D., Beresbaranji, l., Macet, K., Tesovic, Z., Krstic, M., Petrovic, N., 1995. Anatomical and biochemical studies of fruit development in peach. Sci. Hort. 64, 33 48. O'Leary, M.H., 1988. Carbon isotopes in photosynthesis: fractionation techniques may reveal new aspects of carbon dynamics in plants. BioScience 38, 328 336. Pate, J.S., 1975. Exchange of solutes between phloem and xylem circulation in the whole plant. In: Zimmermann, M.H., Milburn, J.M. (Eds.), Encyclopedia of Plant Physiology, New Series, Vol. I: Transport in Plants. I. Phloem Transport. Springer, New York, pp. 451 473. Pate, J.S., Hocking, P.J., 1978. Phloem and xylem transport in the supply of minerals to a developing legume ( Lupinus albus L.) fruit. Ann. Bot. 42, 911 921. Patrick, J.W., 1990. Sieve element unloading: cellular pathway, mechanism and control. Physiol. Plant 78, 298 308. Pitman, M.G., 1975. Whole plants. In: Baker, D.A., Hall, J.L. (Eds.), Transport in Plant Cells and Tissues. Elsevier NorthHolland, Amsterdam, pp. 267- 308. Plummet, A.P., 1977. Revegetation of disturbed lntermountain area sites. In: Thames, J.C. (Ed.), Reclamation and Use of Disturbed Lands of the Southwest. University of Arizona Press, Tuscon, AZ, pp. 302 337. Sale. P.W.G., Campbell, L.C., 1980. Patterns of mineral nutrient accumulation in soybean seed. Field Crops Res. 3, 157 163. Simkiss, K., Wilbur, K.M., 1989. Biomineralization: Cell Biology and Mineral Deposition. Academic Press, Harcourt Brace Jovanovich Publishers, New York, NY, 327 pp. Thomasson, J.R., 1991. Sediment-borne seeds from Sand Creek, northwestern Kansas: taphonomic significance and paleoecological and paleoenvironmental implications. Paleogeogr. Paleoclimatol. Paleoecol. 85, 213 -225. Thorne, J.H., 1985. Phloem unloading of C and N assimilates in developing seeds. Annu. Rev. Plant Physiol. 36, 317 343. Urey, H.C., 1947. The thermodynamic properties of isotopic substances. J. Chem. Soc. 1947, 562 581. van Bel, A.J.E., 1990. Xylem phloem exchange via the rays: the undervalued route of transport. J. Exp. Bot. 41,631 644. Williamson, G.J., Coston, C.D., 1989. The relationship among root growth, shoot growth and fruit growth of peach. J. Am. Soc. Hortic. Sci. 114, 180-183. Wang, Y., Jahren, A.H., Amundson, R.G., 1997. Potential for ~4C dating of biogenic carbonate in hackberry (celtis) endocarps. Quat. Res. 47, 337-343.